scholarly journals Bayesian calibration of a stochastic, multiscale agent-based model for predicting in vitro tumor growth

2021 ◽  
Vol 17 (11) ◽  
pp. e1008845
Author(s):  
Ernesto A. B. F. Lima ◽  
Danial Faghihi ◽  
Russell Philley ◽  
Jianchen Yang ◽  
John Virostko ◽  
...  
Keyword(s):  
Experimental Data ◽  
Cancer Cells ◽  
Temporal Evolution ◽  
Individual Cell ◽  
Coarse Grained ◽  
Computational Time ◽  
Model Parameters ◽  
Dead Cell ◽  
Cell Dynamics ◽  
Agent Based

Hybrid multiscale agent-based models (ABMs) are unique in their ability to simulate individual cell interactions and microenvironmental dynamics. Unfortunately, the high computational cost of modeling individual cells, the inherent stochasticity of cell dynamics, and numerous model parameters are fundamental limitations of applying such models to predict tumor dynamics. To overcome these challenges, we have developed a coarse-grained two-scale ABM (cgABM) with a reduced parameter space that allows for an accurate and efficient calibration using a set of time-resolved microscopy measurements of cancer cells grown with different initial conditions. The multiscale model consists of a reaction-diffusion type model capturing the spatio-temporal evolution of glucose and growth factors in the tumor microenvironment (at tissue scale), coupled with a lattice-free ABM to simulate individual cell dynamics (at cellular scale). The experimental data consists of BT474 human breast carcinoma cells initialized with different glucose concentrations and tumor cell confluences. The confluence of live and dead cells was measured every three hours over four days. Given this model, we perform a time-dependent global sensitivity analysis to identify the relative importance of the model parameters. The subsequent cgABM is calibrated within a Bayesian framework to the experimental data to estimate model parameters, which are then used to predict the temporal evolution of the living and dead cell populations. To this end, a moment-based Bayesian inference is proposed to account for the stochasticity of the cgABM while quantifying uncertainties due to limited temporal observational data. The cgABM reduces the computational time of ABM simulations by 93% to 97% while staying within a 3% difference in prediction compared to ABM. Additionally, the cgABM can reliably predict the temporal evolution of breast cancer cells observed by the microscopy data with an average error and standard deviation for live and dead cells being 7.61±2.01 and 5.78±1.13, respectively.

scholarly journals Bayesian calibration of a stochastic, multiscale agent-based model for predicting in vitro tumor growth

2021 ◽  
Author(s):  
Ernesto A. B. F. Lima ◽  
Danial Faghihi ◽  
Russel Philley ◽  
Jianchen Yang ◽  
John Virostko ◽  
...  
Keyword(s):  
Experimental Data ◽  
Cancer Cells ◽  
Temporal Evolution ◽  
Initial Conditions ◽  
Individual Cell ◽  
Computational Cost ◽  
Coarse Grained ◽  
Model Parameters ◽  
Dead Cell ◽  
Agent Based

Hybrid multiscale agent-based models (ABMs) are unique in their ability to simulate individual cell interactions and microenvironmental dynamics. Unfortunately, the high computational cost of modeling individual cells, the inherent stochasticity due to probabilistic phenotypic transitions, and numerous model parameters that are difficult to measure directly are fundamental limitations of applying such models to predict tumor dynamics. To overcome these challenges, we have developed a coarse-grained two-scale ABM (cgABM) calibrated with a set of time-resolved microscopy measurements of cancer cells grown with different initial conditions. The multiscale model consists of a reaction-diffusion type model capturing the spatio-temporal evolution of glucose and growth factors in the tumor microenvironment (at tissue scale), coupled with a lattice-free ABM to simulate individual cell dynamics (at cellular scale). The experimental data consists of BT474 human breast carcinoma cells initialized with different glucose concentrations and tumor cell confluences. The confluence of live and dead cells was measured every three hours over four days.   Given this model and data, we perform a global sensitivity analysis to identify the relative importance of the model parameters. The subsequent cgABM with a reduced parameter space is calibrated within a Bayesian framework to the experimental data to estimate model parameters, which are then used to predict the temporal evolution of the living and dead cell populations. To this end, a moment-based Bayesian inference is proposed to account for the stochasticity of the cgABM while quantifying uncertainties in model parameters and observational data. The results indicate that the cgABM can reliably predict the spatiotemporal evolution of breast cancer cells observed by the microscopy data with an average error and standard deviation for live and dead cells being 7.61 [[EQUATION]] 2.01 and 5.78 [[EQUATION]] 1.13, respectively.

scholarly journals Speed modulated social influence in evacuating pedestrian crowds

2020 ◽  
Vol 5 ◽  
Author(s):  
Hye Rin Lindsay Lee ◽  
Abhishek Bhatia ◽  
Jenny Brynjarsdóttir ◽  
Nicole Abaid ◽  
Alethea Barbaro ◽  
...  
Keyword(s):  
Experimental Data ◽  
Collision Avoidance ◽  
Social Factors ◽  
Model Parameters ◽  
Force Model ◽  
Confined Space ◽  
Social Force ◽  
Social Force Model ◽  
Agent Based ◽  
Deviation Rate

Evacuation is a complex social phenomenon with individuals tending to exit a confined space as soon as possible. Social factors that influence an individual include collision avoidance and conformity with others with respect to the tendency to exit. While collision avoidance has been heavily focused on by the agent-based models used frequently to simulate evacuation scenarios, these models typically assume that all agents have an equal desire to exit the scene in a given situation. It is more likely that, out of those who are exiting, some are patient while others seek to exit as soon as possible. Here, we experimentally investigate the effect of different proportions of patient (no-rush) versus impatient (rush) individuals in an evacuating crowd of up to 24 people. Our results show that a) average speed changes significantly for individuals who otherwise tended to rush (or not rush) with both type of individuals speeding up in the presence of the other; and b) deviation rate, defined as the amount of turning, changes significantly for the rush individuals in the presence of no-rush individuals. We then seek to replicate this effect with Helbing's social force model with the twin purposes of analyzing how well the model fits experimental data, and explaining the differences in speed in terms of model parameters. We find that we must change the interaction parameters for both rush and no-rush agents depending on the condition that we are modeling in order to fit the model to the experimental data.

scholarly journals An Agent-Based Model of monocyte differentiation into Tumour-Associated Macrophages in Chronic Lymphocytic Leukemia

2021 ◽  
Author(s):  
Nina Verstraete ◽  
Malvina Marku ◽  
Marcin Domagala ◽  
Julie Bordenave ◽  
H&eacutelène Arduin ◽  
...  
Keyword(s):  
Chronic Lymphocytic Leukemia ◽  
Cancer Cells ◽  
Mononuclear Cells ◽  
Lymphocytic Leukemia ◽  
Leukemic Cells ◽  
Patient Specific ◽  
Model Parameters ◽  
Agent Based Model ◽  
Agent Based ◽  
Monocyte Differentiation

Monocyte-derived macrophages are immune cells which help maintain tissue homeostasis and defend the organism against pathogens. In solid tumours, recent studies have uncovered complex macrophage populations, among which tumour-associated macrophages, supporting tumorigenesis through multiple cancer hallmarks such as immunosuppression, angiogenesis or matrix remodelling. In the case of chronic lymphocytic leukemia, these macrophages are known as nurse-like cells and have been shown to protect leukemic cells from spontaneous apoptosis and contribute to their chemoresistance. We propose an agent-based model of monocytes differentiation into nurse-like cells upon contact with leukemic B cells in-vitro. We studied monocyte differentiation and cancer cells survival dynamics depending on diverse hypotheses on monocytes and cancer cells relative proportions, sensitivity to their surrounding environment and cell-cell interactions. Peripheral blood mononuclear cells from patients were cultured and monitored during 13 days to calibrate the model parameters, such as phagocytosis efficiency, death rates or protective effect from the nurse-like cells. Our model is able to reproduce experimental results and predict cancer cells survival dynamics in a patient-specific manner. Our results shed light on important factors at play in cancer cells survival, highlighting a potentially important role of phagocytosis.

A methodology to parameterize SAFT-type equations of state for solid precursors of deep eutectic solvents: the example of cholinium chloride

2019 ◽  
Vol 21 (27) ◽  
pp. 15046-15061 ◽  
Author(s):  
Emanuel A. Crespo ◽  
Liliana P. Silva ◽  
Joel O. Lloret ◽  
Pedro J. Carvalho ◽  
Lourdes F. Vega ◽  
...  
Keyword(s):  
Experimental Data ◽  
Association Scheme ◽  
Equations Of State ◽  
Deep Eutectic Solvents ◽  
Coarse Grained ◽  
Model Parameters

Novel methodology for the development of coarse-grained models applicable to DES – a more realistic association scheme and model parameters regression from experimental data.

scholarly journals Frustration-induced phases in migrating cell clusters

2018 ◽  
Vol 4 (9) ◽  
pp. eaar8483 ◽  
Author(s):  
Katherine Copenhagen ◽  
Gema Malet-Engra ◽  
Weimiao Yu ◽  
Giorgio Scita ◽  
Nir Gov ◽  
...  
Keyword(s):  
Experimental Data ◽  
Velocity Field ◽  
Cancer Cells ◽  
Collective Motion ◽  
Topological Defects ◽  
Local Environment ◽  
Two Dimensional ◽  
Agent Based ◽  
Heterogeneous Behavior

Certain malignant cancer cells form clusters in a chemoattractant gradient, which can spontaneously show three different phases of motion: translational, rotational, and random. Guided by our experiments on the motion of two-dimensional clusters in vitro, we developed an agent-based model in which the cells form a cohesive cluster due to attractive and alignment interactions. We find that when cells at the cluster rim are more motile, all three phases of motion coexist, in agreement with our observations. Using the model, we show that the transitions between different phases are driven by competition between an ordered rim and a disordered core accompanied by the creation and annihilation of topological defects in the velocity field. The model makes specific predictions, which we verify with our experimental data. Our results suggest that heterogeneous behavior of individuals, based on local environment, can lead to novel, experimentally observed phases of collective motion.

scholarly journals Modeling of Disordered Protein Structures Using Monte Carlo Simulations and Knowledge-Based Statistical Force Fields

2018 ◽  
Author(s):  
Maciej Pawel Ciemny ◽  
Aleksandra Elzbieta Badaczewska-Dawid ◽  
Monika Pikuzinska ◽  
Andrzej Kolinski ◽  
Sebastian Kmiecik
Keyword(s):  
Experimental Data ◽  
Monte Carlo ◽  
Protein Folding ◽  
Case Studies ◽  
Protein Structures ◽  
Coarse Grained ◽  
Computational Time ◽  
Modeling Tools ◽  
Disordered Protein ◽  
Knowledge Based

The description of protein disordered states is important for understanding protein folding mechanisms and their functions. In this short review, we briefly describe a simulation approach to modeling disordered protein interactions and unfolded states of globular proteins. It is based on the CABS coarse-grained protein model that uses a Monte Carlo (MC) sampling scheme and a knowledge-based statistical force field. We review several case studies showing that description of protein disordered states resulting from CABS simulations is consistent with experimental data. The case studies comprise investigations of protein-peptide binding and protein folding processes. The CABS model has been recently made available as the simulation engine of multiscale modeling tools enabling studies of protein-peptide docking and protein flexibility. Those tools offer customization of the modeling process, driving the conformational search using distance restraints, reconstruction of selected models to all-atom resolution and studies of large protein systems in a reasonable computational time. Therefore, CABS can be combined in integrative modeling pipelines incorporating experimental data and other modeling tools of various resolution.

scholarly journals Modeling of Disordered Protein Structures Using Monte Carlo Simulations and Knowledge-Based Statistical Force Fields

2019 ◽  
Vol 20 (3) ◽  
pp. 606 ◽  
Author(s):  
Maciej Ciemny ◽  
Aleksandra Badaczewska-Dawid ◽  
Monika Pikuzinska ◽  
Andrzej Kolinski ◽  
Sebastian Kmiecik
Keyword(s):  
Experimental Data ◽  
Monte Carlo ◽  
Protein Folding ◽  
Case Studies ◽  
Protein Structures ◽  
Coarse Grained ◽  
Computational Time ◽  
Modeling Tools ◽  
Disordered Protein ◽  
Knowledge Based

The description of protein disordered states is important for understanding protein folding mechanisms and their functions. In this short review, we briefly describe a simulation approach to modeling protein interactions, which involve disordered peptide partners or intrinsically disordered protein regions, and unfolded states of globular proteins. It is based on the CABS coarse-grained protein model that uses a Monte Carlo (MC) sampling scheme and a knowledge-based statistical force field. We review several case studies showing that description of protein disordered states resulting from CABS simulations is consistent with experimental data. The case studies comprise investigations of protein–peptide binding and protein folding processes. The CABS model has been recently made available as the simulation engine of multiscale modeling tools enabling studies of protein–peptide docking and protein flexibility. Those tools offer customization of the modeling process, driving the conformational search using distance restraints, reconstruction of selected models to all-atom resolution, and simulation of large protein systems in a reasonable computational time. Therefore, CABS can be combined in integrative modeling pipelines incorporating experimental data and other modeling tools of various resolution.

Degrading Bouc–Wen Model Parameters Identification Under Cyclic Load

2017 ◽  
Vol 8 (2) ◽  
pp. 60-81 ◽  
Author(s):  
G. C. Marano ◽  
M. Pelliciari ◽  
T. Cuoghi ◽  
B. Briseghella ◽  
D. Lavorato ◽  
...  
Keyword(s):  
Experimental Data ◽  
Reinforced Concrete ◽  
Structural Engineering ◽  
Bridge Pier ◽  
Computational Time ◽  
Model Parameters ◽  
Concrete Bridge ◽  
Computational Point ◽  
Reinforced Concrete Bridge ◽  
Wide Range

The purpose of this article is to describe the Bouc–Wen model of hysteresis for structural engineering which is used to describe a wide range of nonlinear hysteretic systems, as a consequence of its capability to produce a variety of hysteretic patterns. This article focuses on the application of the Bouc–Wen model to predict the hysteretic behaviour of reinforced concrete bridge piers. The purpose is to identify the optimal values of the parameters so that the output of the model matches as well as possible the experimental data. Two repaired, retrofitted and reinforced concrete bridge pier specimens (in a 1:6 scale of a real bridge pier) are tested in a laboratory and used for experiments in this article. An identification of Bouc–Wen model's parameters is performed using the force–displacement experimental data obtained after cyclic loading tests on these two specimens. The original model involves many parameters and complex pinching and degrading functions. This makes the identification solution unmanageable and with numerical problems. Furthermore, from a computational point of view, the identification takes too much time. The novelty of this work is the proposal of a simplification of the model allowed by simpler pinching and degrading functions and the reduction of the number of parameters. The latter innovation is effective in reducing computational efforts and is performed after a deep study of the mechanical effects of each parameter on the pier response. This simplified model is implemented in a MATLAB code and the numerical results are well fit to the experimental results and are reliable in terms of manageability, stability, and computational time.

scholarly journals Integrating experimental data to calibrate quantitative cancer models

2015 ◽  
Author(s):  
Heiko Enderling
Keyword(s):  
Differential Equation ◽  
Experimental Data ◽  
Cell Cycle ◽  
Cell Proliferation ◽  
Cancer Cells ◽  
Unknown Parameters ◽  
Agent Based ◽  
Cancer Models ◽  
Differential Equation Models ◽  
Model Dynamics

For quantitative cancer models to be meaningful and interpretable the number of unknown parameters must be kept minimal. Experimental data can be utilized to calibrate model dynamics rates or rate constants. Proper integration of experimental data, however, depends on the chosen theoretical framework. Using live imaging of cell proliferation as an example, we show how to derive cell cycle distributions in agent-based models and averaged proliferation rates in differential equation models. We focus on a tumor hierarchy of cancer stem and progenitor non-stem cancer cells.

Parameter Determination for Chloride and Tritium Transport in Undisturbed Lysimeters during Steady Flow

1992 ◽  
Vol 23 (2) ◽  
pp. 89-104 ◽  
Author(s):  
Ole H. Jacobsen ◽  
Feike J. Leij ◽  
Martinus Th. van Genuchten
Keyword(s):  
Experimental Data ◽  
Unsaturated Flow ◽  
Transport Model ◽  
Time Moment ◽  
Breakthrough Curves ◽  
Coarse Sand ◽  
Retardation Factor ◽  
Parameter Determination ◽  
Model Parameters ◽  
Water Fraction

Breakthrough curves of Cl and 3H2O were obtained during steady unsaturated flow in five lysimeters containing an undisturbed coarse sand (Orthic Haplohumod). The experimental data were analyzed in terms of the classical two-parameter convection-dispersion equation and a four-parameter two-region type physical nonequilibrium solute transport model. Model parameters were obtained by both curve fitting and time moment analysis. The four-parameter model provided a much better fit to the data for three soil columns, but performed only slightly better for the two remaining columns. The retardation factor for Cl was about 10 % less than for 3H2O, indicating some anion exclusion. For the four-parameter model the average immobile water fraction was 0.14 and the Peclet numbers of the mobile region varied between 50 and 200. Time moments analysis proved to be a useful tool for quantifying the break through curve (BTC) although the moments were found to be sensitive to experimental scattering in the measured data at larger times. Also, fitted parameters described the experimental data better than moment generated parameter values.

Sign in / Sign up

Export Citation Format

Share Document